Detecting and Preventing Child Cyberbullying using Generative Artificial Intelligence
Pranav Kumar Chaudhary, Sreedhar Yalamati, Naga Ramesh Palakurti, Naved Alam, Saydulu Kolasani, Pawan Whig
Abstract
This research presents a pioneering approach to combat child cyberbullying utilizing generative artificial intelligence (AI) techniques. Our system achieved an impressive detection accuracy of 92.5%, with a precision of 89% and recall of 95%, surpassing traditional methods by 15% in accuracy and 10% in recall. Additionally, the system exhibited rapid response times, with an average of 0.5 seconds for flagging and classifying cyberbullying incidents. Proactive prevention strategies, including the generation of counter-narratives and positive interventions, resulted in a 30% reduction in the escalation of cyberbullying incidents and a 25% increase in the utilization of support resources by affected individuals. Furthermore, the system demonstrated scalability and adaptability, maintaining consistent performance across diverse datasets and showing a 5% increase in detection accuracy over a six-month period. These findings underscore the potential of generative AI in creating safer online environments for children by effectively detecting and preventing cyberbullying behaviors.